diff options
author | Richard Henderson | 2022-06-27 13:17:39 +0200 |
---|---|---|
committer | Richard Henderson | 2022-06-27 13:17:39 +0200 |
commit | 29f6db75667f44f3f01ba5037dacaf9ebd9328da (patch) | |
tree | a0bd532283b2a545635bae5cc1d6affbfad6be45 /docs/system/arm/cpu-features.rst | |
parent | Merge tag 'qemu-sparc-20220626' of https://github.com/mcayland/qemu into staging (diff) | |
parent | target/arm: Check V7VE as well as LPAE in arm_pamax (diff) | |
download | qemu-29f6db75667f44f3f01ba5037dacaf9ebd9328da.tar.gz qemu-29f6db75667f44f3f01ba5037dacaf9ebd9328da.tar.xz qemu-29f6db75667f44f3f01ba5037dacaf9ebd9328da.zip |
Merge tag 'pull-target-arm-20220627' of https://git.linaro.org/people/pmaydell/qemu-arm into staging
target-arm queue:
* sphinx: change default language to 'en'
* Diagnose attempts to emulate EL3 in hvf as well as kvm
* More SME groundwork patches
* virt: Fix calculation of physical address space size
for v7VE CPUs (eg cortex-a15)
# -----BEGIN PGP SIGNATURE-----
#
# iQJNBAABCAA3FiEE4aXFk81BneKOgxXPPCUl7RQ2DN4FAmK5hKEZHHBldGVyLm1h
# eWRlbGxAbGluYXJvLm9yZwAKCRA8JSXtFDYM3lggEACtE2balVHrVCeSQkRW+FnS
# avm5i54MIGf3cgNhTKwD9ED3hl03Xm49yQkaq0gB6Qa4wQPEcYQLSyzP+UYIILO5
# 3xoWEw0nbtKWBuCzdiolynL1VFht6GV+Ga8lShoBiQsI/eARC6ZELvBv7gbApf4p
# DpDq1ty7fXMmMCNM5vgX9fu/LXahSONDXbYMpHpohnaLXCEF9MwqpO5TJf65Bgze
# z2+NO4R5u26mCcad7ltoiz3OKkq4Bq+b+QXrm6LmvSCIkvk6MUZuU1NwHSiqUoV/
# nOwhJriOVl8JG0sX0xzNZADYBt0YlcVuDZzyxP8eOiQ54CVK7rJOJSi+aiGkg2Mn
# YC4CkFZY9iM5YTA6y6T5mye7kLb/pJ746rLM1ia6Ng3rUwoE9bdvruqTMfPPJuoo
# XxMBQrjRjY6BzESG0NbjLgg80dPtqeOipjglYI7GCvh0i2yQVmKLQon5TK9DsScC
# 7Gu6IPVWZAb3axGEuqjJ4E+7PyyEW7zYgWNOpZoQW958WHDK0KSPrOwqxAC+QdEi
# vagKJGCQPuZiOARpXm6F/nscEDcy7P33z120O9/R6HuticGaM/oBaWy89CR4hbHB
# NWx5+0h5M/je8hJFJJVfHldR3nIpvnUtb4KEVoiNuxkrGZoejgTlBdKNL4Nph0U0
# E+CQyMuBBQ88LEbyCjJS5w==
# =GILG
# -----END PGP SIGNATURE-----
# gpg: Signature made Mon 27 Jun 2022 03:51:21 PM +0530
# gpg: using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg: issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [full]
# gpg: aka "Peter Maydell <pmaydell@gmail.com>" [full]
# gpg: aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [full]
* tag 'pull-target-arm-20220627' of https://git.linaro.org/people/pmaydell/qemu-arm: (25 commits)
target/arm: Check V7VE as well as LPAE in arm_pamax
target/arm: Extend arm_pamax to more than aarch64
target/arm: Move pred_{full, gvec}_reg_{offset, size} to translate-a64.h
target/arm: Add SVL to TB flags
target/arm: Introduce sve_vqm1_for_el_sm
target/arm: Add cpu properties for SME
target/arm: Unexport aarch64_add_*_properties
target/arm: Move arm_cpu_*_finalize to internals.h
target/arm: Generalize cpu_arm_{get, set}_default_vec_len
target/arm: Generalize cpu_arm_{get,set}_vq
target/arm: Create ARMVQMap
target/arm: Move error for sve%d property to arm_cpu_sve_finalize
target/arm: Implement SMSTART, SMSTOP
target/arm: Add the SME ZA storage to CPUARMState
target/arm: Add PSTATE.{SM,ZA} to TB flags
target/arm: Add SMIDR_EL1, SMPRI_EL1, SMPRIMAP_EL2
target/arm: Add SMCR_ELx
target/arm: Add SVCR
target/arm: Add ARM_CP_SME
target/arm: Add syn_smetrap
...
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Diffstat (limited to 'docs/system/arm/cpu-features.rst')
-rw-r--r-- | docs/system/arm/cpu-features.rst | 56 |
1 files changed, 56 insertions, 0 deletions
diff --git a/docs/system/arm/cpu-features.rst b/docs/system/arm/cpu-features.rst index 3e626c4b68..3fd76fa0b4 100644 --- a/docs/system/arm/cpu-features.rst +++ b/docs/system/arm/cpu-features.rst @@ -372,6 +372,31 @@ verbose command lines. However, the recommended way to select vector lengths is to explicitly enable each desired length. Therefore only example's (1), (4), and (6) exhibit recommended uses of the properties. +SME CPU Property Examples +------------------------- + + 1) Disable SME:: + + $ qemu-system-aarch64 -M virt -cpu max,sme=off + + 2) Implicitly enable all vector lengths for the ``max`` CPU type:: + + $ qemu-system-aarch64 -M virt -cpu max + + 3) Only enable the 256-bit vector length:: + + $ qemu-system-aarch64 -M virt -cpu max,sme256=on + + 3) Enable the 256-bit and 1024-bit vector lengths:: + + $ qemu-system-aarch64 -M virt -cpu max,sme256=on,sme1024=on + + 4) Disable the 512-bit vector length. This results in all the other + lengths supported by ``max`` defaulting to enabled + (128, 256, 1024 and 2048):: + + $ qemu-system-aarch64 -M virt -cpu max,sve512=off + SVE User-mode Default Vector Length Property -------------------------------------------- @@ -387,3 +412,34 @@ length supported by QEMU is 256. If this property is set to ``-1`` then the default vector length is set to the maximum possible length. + +SME CPU Properties +================== + +The SME CPU properties are much like the SVE properties: ``sme`` is +used to enable or disable the entire SME feature, and ``sme<N>`` is +used to enable or disable specific vector lengths. Finally, +``sme_fa64`` is used to enable or disable ``FEAT_SME_FA64``, which +allows execution of the "full a64" instruction set while Streaming +SVE mode is enabled. + +SME is not supported by KVM at this time. + +At least one vector length must be enabled when ``sme`` is enabled, +and all vector lengths must be powers of 2. The maximum vector +length supported by qemu is 2048 bits. Otherwise, there are no +additional constraints on the set of vector lengths supported by SME. + +SME User-mode Default Vector Length Property +-------------------------------------------- + +For qemu-aarch64, the cpu propery ``sme-default-vector-length=N`` is +defined to mirror the Linux kernel parameter file +``/proc/sys/abi/sme_default_vector_length``. The default length, ``N``, +is in units of bytes and must be between 16 and 8192. +If not specified, the default vector length is 32. + +As with ``sve-default-vector-length``, if the default length is larger +than the maximum vector length enabled, the actual vector length will +be reduced. If this property is set to ``-1`` then the default vector +length is set to the maximum possible length. |